Hey, it looks like you shut it off entirely (a few seconds ago?). For what it's worth, I didn't hate it. If it's a necessary performance change, people will understand. It was nice to see HN trying an experiment; I'm sorry if my comments sounded negative.
From analyzing it today, one alarming thing I noticed was that in all recent threads with "more" links, the conversation never seemed to nest more than one or two levels deep. But the moment you go to an old thread, e.g. https://news.ycombinator.com/item?id=21231208#21232605, the "more" links turned out to be hiding 9+ comments.
But, like, it's not a horrible thing to constrain the comment subtree. If you were feeling like it was the way to go, my only feedback is to trust your judgement on that one. Personally, I thought it was an interesting new dynamic, even if I had my own (probably-unjustified) concerns.
Anyway, best of luck, sorry for the mistaken bug report, and keep trying new experiments! It's really cool!
Honestly, at HN’s scale, maybe it makes sense now. After spending today using it, my gut reaction was to be sad that people will feel reluctant to throw a bunch of effort into replies that will immediately be hidden (as with Reddit). But after the server meltdown yesterday, it’s understandable.
14 years was a really good run. I suppose even lisp has its limits. :(
Yeah, adding the descendant count would make it pretty good. Hard to think of anything better.
I’ll be honest: I’m really worried this will become another tool for moderators to control discussions, while concealing the fact that it was a moderator (rather than algorithmic) decision. But no one can stop you if that’s the kind of control you want for your team.
The flip side of that is, since 2018, your team has done a stellar job, as far as I can tell. I don’t know if something changed, but things seem calmer now. I can’t even recall the last worrisome moderator decision, and I’ve found myself arguing on your behalf more than once.
Dunno. You’ll get it right; you usually do. Thanks for being welcoming about ideas, and good luck wrestling with transpilers and performance issues.
We're experimenting, in order to deal with this problem: https://news.ycombinator.com/item?id=27437362.